Abstract

In this paper I examine Pierre Duhem's position on the problem of crucial experiments and the falsification of hypotheses. Duhem maintained that conclusive falsification of an isolated hypothesis is impossible; he maintained further that crucial experiments are impossible. But, I argue, this does not imply, as Adolf Grunbaum and others have taken Duhem to imply, that one could always save any hypothesis by making adjustments elsewhere in the system. By analyzing the logic of falsification, by examining the text of Duhem's The Aim and Structure of Physical Theory, and by surveying the recent articles about Duhem I show the correct interpretation of Duhem, the primary reason for the misinterpretations of him, and the value of Duhem's position.